wordcluster

library(factoextra)

## Loading required package: ggplot2

## Welcome! Related Books: `Practical Guide To Cluster Analysis in R` at https://goo.gl/13EFCZ

library(plotly)

## 
## Attaching package: 'plotly'

## The following object is masked from 'package:ggplot2':
## 
##     last_plot

## The following object is masked from 'package:stats':
## 
##     filter

## The following object is masked from 'package:graphics':
## 
##     layout

#######
d <- read.csv("dfall.csv", sep = ",") 

all.data <- d[,-c(1,2)] #delete the numbering column

rownames(all.data) <- d$.rownames
all.data <- all.data[-which(d$.rownames %in% c("and","with","for","when","the")),]

colnames(all.data) <- c("P1","P2","P3","P4","P5","P6","P7","P8","P9","P10","P11","P12","P13","P14","P15","P16","P17","P18","P19","P20","P21","P22","P23","P24","P25","P26","P27","P28",
                          "P29","P30","P31","P32","P33","P34","P35","P36","P37")


#### sort patients by the length they lived(days) 
c1<- c("P12", "P23", "P13", "P3", "P5", "P7", "P21", "P16", "P17")
c11 <- c("P37", "P31", "P2", "P11")
c21 <- c("P29", "P6", "P26", "P10", "P9")
c31 <- c("P28", "P19", "P1", "P18", "P4")

# Elbow Method


library(dplyr)

## 
## Attaching package: 'dplyr'

## The following objects are masked from 'package:stats':
## 
##     filter, lag

## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

###c1, rid of rows that has 0 as sum

vis <- function(x,y){
  sel <- all.data %>% select(x) 
  sum<- rowSums(sel)
  sel <- cbind(sel, sum)
  no0 <-sel[sum !=0,]
  run<- no0[,-10] 
  
  set.seed(20)
  km <- kmeans(run, centers = 3, nstart = 20)
  return(ggplotly(fviz_cluster(km, data = run, main = y)))
}

gp1 <- vis(c1, "word cluster(day=1~10)")

gp2 <- vis(c11, "word cluster(day=11~20)")

gp3 <- vis(c21, "word cluster(day=21~30)")

gp4 <- vis(c31, "word cluster(day=31~)")

gp1

gp2

gp3

gp4